Integrated Campaign for Aerosols, gases and Radiation Budget (ICARB): An overview
K Krishna Moorthy1,S K Satheesh2,S Suresh Babu1and C B S Dutt3
1Space Physics Laboratory, Vikram Sarabhai Space Centre, Trivandrum 695 022, India.
2Centre for Atmospheric and Oceanic Sciences, Indian Institute of Science, Bangalore 560 012, India.
3Indian Space Research Organisation Head Quarters, New BEL Road, Bangalore 560 094, India.
During March–May 2006, an extensive, multi-institution, multi-instrument, and multi-platform integrated field experiment ‘Integrated Campaign for Aerosols, gases and Radiation Budget’
(ICARB) was carried out under the Geosphere Biosphere Programme of the Indian Space Research Organization (ISRO-GBP). The objective of this largest and most exhaustive field campaign, ever conducted in the Indian region, was to characterize the physico-chemical properties and radiative effects of atmospheric aerosols and trace gases over the Indian landmass and the adjoining oceanic regions of the Arabian Sea, northern Indian Ocean, and Bay of Bengal through intensive, simul- taneous observations. A network of ground-based observatories (over the mainland and islands), a dedicated ship cruise over the oceanic regions using a fully equipped research vessel, the Sagar Kanya, and altitude profiling over selected regions using an instrumented aircraft and balloonson- des formed the three segments of this integrated experiment, which were carried out in tandem.
This paper presents an overview of the ICARB field experiment, the database generated, and some of its interesting outcomes though these are preliminary in nature.
The ICARB has revealed significant spatio-temporal heterogeneity in most of the aerosol char- acteristics both over land and ocean. Observed aerosol loading and optical depths were comparable to or in certain regions, a little lower than those reported in some of the earlier campaigns for these regions. The preliminary results indicate:
• low (<0.2) aerosol optical depths (AOD) over most part of the Arabian Sea, except two pockets; one off Mangalore and the other, less intense, in the central Arabian Sea at
∼18◦N latitude;
• High ˚Angstr¨om exponent in the southern Arabian Sea signifying steep AOD spectra and higher abundance of accumulation mode particles in the southern Arabian Sea and off Mangalore;
• Remarkably low ˚Angstr¨om exponents signifying increased concentration of coarse mode aerosols and high columnar abundance in the northern Arabian Sea;
• Altitude profiles from aircraft showed a steady BC level up to 3 km altitude with structures which were associated with inversions in the atmospheric boundary layer (ABL);
• A surprisingly large increase in the BC mass fraction with altitude;
• Presence of a convectively mixed layer extending up to about 1 km over the Arabian Sea and Bay of Bengal;
• A spatial off shore extent of<100 km for the anthropogenic impact at the coast; and
• Advection of aerosols, through airmass trajectories, from west Asia and NW arid regions of India leading to formation of elevated aerosol layers extending as far as 400 km off the east coast.
Keywords. ICARB, atmospheric aerosols; radiative forcing; Bay of Bengal and Arabian Sea aerosols; altitude profiles;
Micro Pulse Lidar.
J. Earth Syst. Sci.117, S1, July 2008, pp. 243–262
©Printed in India. 243
1. Introduction
Atmospheric aerosols play an important role in the Earth’s energy budget and contribute to climate change. They affect the climate directly by scat- tering and absorbing radiation (McCormick and Ludwig 1967; Charlson et al 1992) and indirectly by acting as cloud condensation nuclei and thereby affecting cloud properties (Twomey 1977). How- ever, radiative forcing and the resulting climate impact due to atmospheric aerosols remain largely uncertain primarily due to inadequate data rep- resenting the spatio-temporal heterogeneity of the aerosol properties and in the understanding of aerosol cloud interactions (IPCC 2007). The uncer- tainty in the aerosol direct climate forcing is esti- mated to be in the range of −0.1 to −0.9 W m−2 (with a mean value of −0.5 W m−2), while it is much higher for the indirect forcing (varying from −0.3 to −1.8 W m−2) with a mean value of
−0.7 W m−2. Reduction of these uncertainties calls for a co-ordinated research strategy that will suc- cessfully integrate data from multiple platforms such as ground-based network, ship, aircraft, bal- loon, and satellite, as well as numerical modeling and data assimilation (Penneret al 1994; Kaufman et al 2002; Diner et al 2004; Yu et al 2006). This is particularly needed for the Asian region, with all its natural diversities, high density of population, diverse living habits, and the growing industriali- zation and urbanization.
Over the past decade, more than a dozen intense field experiments have been conducted to characterize physical, chemical, and optical prop- erties, and the radiative effects of aerosols in a variety of aerosol regimes around the world.
These include BASE-A over Brazil (Kaufmanet al 1992), SAFARI-92 over the southern Atlantic and southern Africa (Lindesay et al 1996), TRACE-A over the southern Atlantic (Fishman et al 1996), SCAR-A over North America (Remer et al 1997), SCAR-B over Brazil (Kaufmanet al 1998), ACE-1 over the Southern Oceans (Bates et al 1998), TARFOX over the northwestern Atlantic (Russell et al 1999), ACE-2 over the northern Atlantic (Raes et al 2000), Aerosols99 over the Atlantic (Bates et al 2001), INDOEX over the Indian Ocean (Ramanathan et al 2001), SAFARI-2000 over South Africa and the south Atlantic (King et al 2003), PRIDE over Puerto Rico (Reid et al 2003), TRACE-P over east Asia and the north- western Pacific (Jacob et al 2003), ACE-Asia over east Asia and the northwest Pacific (Huebert et al 2003), MINOS over the Mediterranean region (Lelieveld et al 2002), CLAMS over the east coast of U.S. (Smith et al 2005) and NEAQS over the north Atlantic (Quinn and Bates 2005). Dur- ing each of these comprehensive missions, aerosols
were studied in great detail, using combinations of in situ and remote sensing observations of physical and chemical properties from different platforms. In addition, long term aerosol moni- toring network such as AERONET of NASA (Holben et al 1998), sunphotometer network in east Asia (SKYNET, http://atmos.cr.chiba- u.ac.jp/aerosol/skynet), multifilter shadowband radiometer (MFRSR) network in north America (Alaxandrov et al 2002), Micropulse Lidar Net- work (MPLNET) of NASA (Welton et al 2001), and European Aerosol Research Lidar Network (Matthias et al 2004) were pursued, essentially to understand the complex physical and optical pro- perties of atmospheric aerosols.
Despite the large diversities in aerosol char- acteristics over India, a concerted effort leading to development of region-specific and season- specific aerosol models remains to be achieved.
The Geosphere Biosphere Programme of the Indian Space Research Organization (ISRO–GBP) how- ever, has been addressing this problem in a focused manner using a chain of ground-based observato- ries over the mainland and islands in the Arabian Sea (AS) and Bay of Bengal (BoB) (now called the ARFI network, ARFI standing for Aerosol Radiative Forcing over India), supplemented with pilot campaigns (for e.g., Moorthy et al 2005a, b;
Gangulyet al 2005; Vinojet al 2004; Tripathiet al 2005; Pant et al 2006; Babu et al 2002, 2004).
Based on the inferences drawn from such efforts, an integrated campaign, the first of its kind over this region, was conceived and executed during the March–May period of 2006. This campaign, ICARB (Integrated Campaign for Aerosols, gases, and Radiation Budget), is also the first of a series of such campaigns planned under ISRO-GBP. Brief descriptions of the campaign concept and details of its execution are given in the following sections with some preliminary results.
2. Scientific objectives and measurements
The ICARB essentially addressed the following issues:
• What is the nature of the spatio-temporal het- erogeneities of the aerosol environment over India and the adjoining oceans? Where are the hotspots (natural/anthropogenic) and sinks?
How does the spatial distribution change with time?
• What is the nature of the vertical distribution of aerosol properties over the coastal regions which interface between the more polluted continental regions and the less polluted oceanic regions?
How strong are the elevated aerosol layers at different regions?
• What are the influences of long-range transport and regional meteorology on the spatial distri- bution of aerosols and gases? How strong are the impacts of advection from different source regions in India, south China, east Asia, etc., over the BoB and India, west Asia, and Africa over the Arabian Sea? How does the change in the prevailing circulation pattern modify the regional distribution?
• What is the net impact of aerosols on regional radiative forcing and how does this forcing change with the above processes?
• What are the impacts of mesoscale processes like the land/sea-breeze activity on the surface con- centrations and altitude profiles of aerosols along the coastal regions? What is the spatial variation of marine atmospheric boundary layer (MABL) over the BoB and Arabian Sea and what are its implications to vertical and horizontal exchanges?
In order to address the above questions, the fol- lowing parameters pertaining to aerosols and gases were measured or estimated using collocated mea- surements during ICARB:
• Spectral Aerosol Optical Depth, AOD (over the land network, onboard the ship and from Mini- coy and Port Blair islands).
• Mass concentration and size distribution of com- posite aerosols over the land and ocean.
• Bulk aerosol sampling from coastal, inland and oceanic environments for identifying and distin- guishing major chemical species as well as to assess the source strength of different geogra- phical regions using the chemical composition as fingerprints.
• Mass concentrations of aerosol black carbon over selected land stations, islands, and onboard the cruise and aircraft.
• Balloon and SODAR measurements of boundary layer parameters to understand the vertical mix- ing in the marine atmospheric boundary layer over the oceans.
• GPS-sonde measurements of the altitude profiles of ozone over the oceanic regions.
• Altitude profiles of aerosol extinction using micro-pulse lidar (MPL) onboard the aircraft supplemented with similar measurements from the adjoining coastal mainland in conjunction with the cruise measurements.
• Aircraft measurements of vertical profiles of important aerosol species and trace gases over the study area and horizontal gradients to under- stand the spatial dispersion of these species and to model them.
3. Period of the campaign
The campaign, spanning for about two months, was conducted in the window March to May 2006.
The main rationales in selecting the campaign window were:
• Availability of uninterrupted sunshine over the entire region (India and adjoining continents as well as over the oceanic region) for most of the time.
• The period forms part of the long, Asian dry sea- son, with very little precipitation and hence the aerosols have long atmospheric residence time even in the lower troposphere.
• The period also favours a significant amount of long-range “trans-boundary transport” of aerosols and trace gases from central India, east and southeast Asia to the BoB and west Asia and east Africa, as well as India to the Arabian Sea (Moorthy et al 2003, 2005a).
4. The structure of ICARB
The campaign was structured following an
‘integrated-segmented’ approach. It consisted of three observation-intense segments; viz., the land segment, the ocean segment, and the air segment.
Each segment was planned to be self-contained and complete, the data from all these will be synthe- sized to characterize the spatio-temporal features, which would then be used for regional radiative forcing.
4.1 The land segment
The land segment comprised fixed aerosol obser- vatories, making continuous time series measure- ments of several aerosol parameters following a common protocol. These observatories are spread over the Indian mainland as shown by the circles encircling the star symbols in figure 1, as well as two island observatories, Minicoy (MCY) in the AS and Port Blair (PBR) in the BoB. Here the observations are carried out on a continuous basis.
In all, there were 18 stations operating in the net- work. The stations and the measurements from these are listed in table 1.
4.2 The ocean segment
The ocean segment of ICARB was designed so as to obtain a snapshot of various parameters of aerosols and gases (that are important in estimat- ing the radiative forcing), using collocated instru- ments, operating as far as possible, under the same environmental conditions, covering the vast
Figure 1. Schematic representation of the operational phase of ICARB during March–May 2006. The continuous lines show the cruise track of SK223 (in the ocean segment). The points on the lines indicate the ship’s location at 05:30 UTC for the date identified below it; the number stands for the day and the alphabet for the month, M for March, A for April, and My for May. The bases from where the aircraft sorties were made are identified with a symbol of the aircraft. The other points represent the fixed stations of the land segment, from where aerosol and gas measurements were made.
oceanic regions of the BoB, northern Indian Ocean (IO), and the AS, during a climatologically dis- tinct season when the prevailing conditions do not change drastically. As such, measurements of all the relevant parameters were carried out onboard the oceanographic research vessel (ORV) Sagar Kanya, of the Department of Ocean Development, the cruise# SK223, which was dedicated to the ICARB, to cover the entire oceanic regions of interest in as short a time-span as possible. This would provide a statistically stationary (in tem- poral scale) spatial mosaic of all the parameters with a reasonable spatial resolution needed for inputting to the current General Circulation Mod- els (GCM) for climate impact assessment. More- over, the climatological winds over these regions and their relevance to transport aerosols to dif- ferent oceanic regions, and the extent of cloudi- ness were also considered. The cruise had two main legs; SK223A covering the BoB and northern IO regions, and 223B covering the Arabian Sea. There was a short cruise for a period of one week, pre- ceding SK223A, which was considered as a dry run, and during this some of the aerosol instru- ments were set up, fine tuned and calibrated. Sev- eral measurements were also initiated in this phase.
Details are given in table 2, while the tracks are shown in figure 1 by the lines over the oceanic regions. Overall the ship sailed for 56 days dur- ing ICARB, covering a track length of >25,000 km
surveying the atmosphere over an oceanic area
>4 million km2. Important urban conglomerations and ports on the mainland, en route the cruise, are marked in figure 1, so that the rationale of the track would be easily understood. During its course, the cruise collected atmospheric data from coastal regions adjacent to major urban centres such as Goa, Mangalore, Kochi, Chennai, Nellore, Vijayawada, Visakhapatnam, Bhubaneswar/Puri, Kolkata, Chittagong, Ranagoon, Port Blair, Colombo (and southern Sri Lanka), Trivandrum, Lakshadweep, Mumbai, Kutch, Kandla and the eastern coasts of Oman, Yemen and Muscat and put these data in juxtaposition with those from far and pristine oceanic environments of the BoB and Arabian Sea.
In the ocean segment collocated measurements of more than 20 parameters of aerosols and trace gases, data on the state of the atmosphere (including the marine atmospheric boundary layer, MABL), and vertical profiles of meteorological parameters and concentration of ozone, were made using about 40 different instruments by a team from 26 national laboratories, academic institu- tions, and universities, onboard the ORV. For aerosols and gas measurements, two dedicated lab- oratories were set up in the top (B and C) decks of the ORV, with community air inlets providing the air samples needed for all the indoor measur- ing instruments (figure 2, top). All the outdoor
Table 1. Ground network stations and measurements.
Station Location Parameters measured Instruments used
Minicoy 8.2◦N, 73.0◦E, 1 m AOD, NSD,MB MWR, MTPS, OPC, Aethalometer
Trivandrum 8.55◦N, 77◦E, 3 m AOD, NSD,MB MWR, ELPI,
Aethalometer
Bangalore 13◦N, 77◦E, 960 m AOD MWR
Anantapur 14.7◦N, 77.6◦E, 331 m AOD MWR
Kalpakkam 12.56◦N, 80.17◦E, 12 m AOD,MB, Scatt coeff MTPS, Aethalometer, Nephelometer
Port Blair 11.63◦N, 92.7◦N, 73 m AOD,MT, MSD,MB MWR, QCM, Aethalometer Visakhapatnam 17.7◦N, 83.3◦E, 5 m AOD, EXTN MWR, MPL
PROFILE
Hyderabad 17.48◦N, 78.4◦E, 545 m AOD,MB MWR, Aethalometer
Pune 18.72◦N, 73.85◦E, 559 m MB Aethalometer
New Delhi 28.6◦N, 77.1◦E, 213 m AOD,MB Microtops,
Aethalometer
Dehra Dun 30.34◦N, 78.04◦E, 690 m AOD MWR
Patiala 30.33◦N, 76.46◦E, 251 m AOD MWR
Kullu 31.9◦N, 77.1◦E, 1155 m AOD MWR
Nainital 29.2◦N, 79.3◦E, 1950 m AOD, NSD,MT,MB MWR, OPC, HVS, Aethalometer
Kanpur 26.4◦N, 80.3◦E, 142 m AOD,MB CIMEL, Aethalometer
Kharagpur 22.3◦N, 87.3◦E, 40 m MB Aethalometer
Dibrugarh 27.3◦N, 94.6◦E, 104 m AOD MWR
Abbreviations: AOD: Aerosol Optical Depth,MB: Mass concentration of BC,MT: Mass concentration of total aerosols and NSD: Number size distribution.
MWR: Multi-Wavelength Radiometer; MTPS: Microtops instrument; OPC: Optical particle counter; HVS:
High volume sampler; CIMEL: Cimel radiometer, QCM: Quartz crystal microbalance impactor, ELPI:
Electrostatic Low Pressure Impactor.
Table 2. Details of the cruise of ORV Sagar Kanya during ICARB.
Days
Cruise leg Period Regions covered sailed
Dry run March 09 to March 13, 2006
Goa to Chennai, along the coastal Arabian Sea, around Sri Lanka and coastal Bay of Bengal
5
SK223 A March 18 to April 13, 2006
Chennai to Kochi, surveying the entire BoB and northern IO and south coastal Arabian Sea
28
SK223 B April 18 to May 11, 2006
Kochi to Goa surveying the entire Arabian Sea
24
sampling instruments were installed at the fore of the B deck (figure 2, bottom) so that they also sample the same ambient air as that by the indoor analysers. These inlets drew air from the port side of the ship, about 1 meter off the ship’s hull and
∼10 m above the water level. The inlets were so configured that the air came on to them as the ship moved so that the sampling was into the wind, devoid of any contamination from the ship.
Besides, the ship was continuously in motion (for
>22 h a day), except for short stoppages for the launching of balloons. Even during this period, the ship was positioned such that the winds arrive from the port side, while the balloons are launched from the starboard side to ensure that the data were always uncontaminated. During the port calls, all the instruments were switched off once the ship entered the berth. A list of major instruments
Figure 2. (a)Photograph of the aerosol and gas laboratory on the B deck of the ORV showing the different instruments connected to the community air inlets.(b)An array of out- door aerosol instruments on the B deck sampling into the incoming air.
operated onboard the ORV during ICARB is given in table 3, along with references to the instrumen- tation and data deduction.
Daily balloon ascents were made from the ship for measuring the altitude profiles of meteoro- logical parameters, using GPS based Vaisala son- des. Ozonesonde ascents were made on alternate days. These provided high-resolution altitude pro- files up to the mid-stratosphere (25 to 35 km).
Other onboard measurements included sun pho- tometer measurements of spectral AOD (aerosol optical depth), column ozone, column water vapour (using sun photometer and GPS), and radiative fluxes in the UV, visible and total short wave. The ORV also carried onboard two automatic weather stations and a Doppler sodar for boundary layer measurements.
4.3 The air segment
The air segment of ICARB was executed using the aircraft of the National Remote Sensing Agency (NRSA), which dedicated one of its propeller air- craft (Beechcraft 200, used mainly for aerial sur- vey). The sorties were carried out from five bases, two each on the east and west coasts of India and one from the interior continent. The sorties from each base were carried out in close co- ordination with the network and cruise opera- tion, so that complementary, near simultaneous surface measurements are ensured around the period and location of the airborne measurements.
The sorties were planned to be as close to the ORV tracks in the adjoining oceanic region as was possible. The sorties were carried out during the daytime and night-time. While the night-time sortie was dedicated for vertical profiling of aerosol
extinction using a micro-pulse lidar mounted (in down-looking mode) over the optical flat win- dow of the aircraft, the daytime sorties made in situ measurements of a number of parameters of aerosols and gases in the lower atmospheric region up to ∼3 km from the ground, the ceiling height of the aircraft permissible in unpressurised condi- tion. In all, 26 sorties were made from five bases, as outlined in figure 3(a). Each station had a suit of sorties consisting of the following:
• One high-resolution profiling of the aerosol para- meters during the forenoon period, after the ABL has evolved. This was aimed at understan- ding structures if any in the profile and the role of the atmospheric dynamics on them. During this, the aircraft climbed the altitudes in a step- and-staircase mode from ground to 3000 m in 8 levels (at 500, 800, 1100, 1400, 1700, 2000, 2500 and 3000 m); but of short horizontal extents.
This profile was made over the oceanic region, adjacent to the base, about 50 km offshore and thus represented the scenario over coastal ocean.
• A longitudinal profile, where the measurements were made across the longitudes, from the base into the oceanic atmosphere over a longitudi- nal span of ∼4◦, to look for gradients over the ocean. Measurements were made only at two alti- tudes, 500 m (well within the ABL) and 1500 m (above the ABL, but below the trade wind inver- sion level), so that appreciably long latitudi- nal/longitudinal coverage is ensured within the available flight window.
• A latitudinal profile, along the coastline as far as possible, following the same criteria discussed for the sortie no 2 above. Two sorties were made in this category to get a wide coverage.
• One night-time sortie, exclusively for the micro- pulse lidar observation, during which the aircraft was pressurized and taken to an altitude of 8 km from where the lidar was operated in a down- looking mode, flying from the land towards the ocean and back.
The list of instruments operated in the air seg- ment and the parameters measured are given in table 4. The ambient air was aspirated into the air- craft using stainless steel inlet pipe, fitted under the nose of the aircraft, such that the inlet opens into the incoming air as the aircraft flies. The oper- ating conditions and flow rate of the instruments were set to ensure sampling of as large a volume as possible within the shortest time permitted, still all conditions were within the permitted limits of each instrument. Wherever needed, the data were corrected for the change in the pumping speed with decrease in the ambient pressure as the aircraft climbed to higher levels (Moorthyet al 2004).
Table3.Instrumentsandmeasurementsonboardtheship. Sl.no.InstrumentReferenceParametersmeasured/retrievedPrinciple 1Microtopssunphotometer andozonemonitorMorysetal(2001)•ColumnarAODat5wave- lengths •ColumnarO3 •Columnarwatervapour •AODat1020nm
Langleytechniquebasedon internalcalibrationconstants. Differentialabsorptionmethod, intheUVandnearIR.Cali- brated 4QCMimpactor–10 channelsPillaiandMoorthy(2001)Masssizedistributionandtotal massconcentrationofcompos- iteaerosolsinthesizerange0.05 to25µm
Peizo-electricmasssensingtech- nique 5Aethalometer–7 channelsHansenetal(1984) BabuandMoorthy(2002)MassconcentrationofBC aerosolsPhotometric–changeintrans- missionaftersampling 6OpticalParticleCounterwww.grimm-aerosol.com Pantetal(2006)Numberconcentrationofcom- positeaerosolsin16channelsin thesizeregime0.3to25µm
Opticalscatteringbysinglepar- ticle 7SequentialMobility ParticleSizer+Counterwww.grimm-aerosol.comNumbersizedistributionofsub- micronsizedcompositeaerosols from10nmto1000nmin42 channels
SizesegregationusingaDMA andopticalcountingofsingle particleusingaCPC 8IntegratingNephelometerAndersonetal(1996)Totalscatteringcross-section andbackscatteringcross- sectionofcompositeaerosols (lessthan1micron)atthree wavelengths
Scatteringoflightbyindividual particlesanddetectionatthree wavelengths 9Highvolumesamplers –Singlestageandmulti stages
Nairetal(2007)•Massconcentrationofbulk andsizesegregatedaerosols •Chemicalcomposition
Collectiononfiltersubstrates, andoff-linechemicalspeciation andgravimetricanalysis 10GasanalysersNajaetal(1999)VolumeconcentrationsofO3, CO,NOxandSO2Photometric–differential absorptionintheUVandIR 11Radiationinstruments Shortwavepyranometer andNetradiationmeter
Satheeshetal(1999)Down-wellingsolarradiation NetradiationAbsolutemeasurementsof globalbroadbandradiations (solarandnet)usingcalibrated sensors–keptonagimbals mount
Table3.(Continued). Sl.no.InstrumentReferenceParametersmeasured/retrievedPrinciple 12DopplerSodar–with beamsteering–oper- atedinmonostaticmode presentlyMonostaticSodar at2.2KHz Crescenti(1997)ThermalstructureandC2 T. Duetohighbackgroundnoise, Dopplerinformationandwinds arenotrecorded.
Backscatteringofsoundbytem- peratureturbulenceandmea- surementofwindsfromthe Dopplershift.(Currentlynor Thermalplumeheight–upto ∼700mBackscatteringofsoundby turbulence 13GPSreceiveranddata loggerContinuousinformationof UTC,ship’sposition(lat., long.,alt.),speed,andheading. Dataupdatedeverysecond
Basicinformation.Alsoused forcorrectingtheship’smotion whileretrievingtruewindsfrom theapparentwinds DifferentialGPSreceiverWolfeetal(2000)Columnarprecipitablewater vapourUseofdualfrequencyconfigu- rationtoestimatetropospheric watervapour 14Vaisala-sondeVerticalprofilesoftemperature, RH,andwindfieldsat10mres- olutionupto25km
Balloonborne,GPSbased radiosondeascentoneperday Ozone-sondeVerticalprofilesofOzone concentration,temperatue, RHandwindfieldsat30mres- olutionupto34km
Balloonborne,GPSbased radiosondeandozonesondeon chemicaltitrationtechnique. Soundingonalternatedays 15Fastresponsesensorson retractableboomBoundarylayerfluxesofsensi- bleandlatentheatsbyprofiling method.
Two-levelmeasurementsof winds,temperatureandRH usingfastresponsesensors mountedonaretractableboom projectingoutwardsatthe ship’sbow 16Meteorologialkit Anemometer,windwane, wetanddrybulbther- mometrs,MarineBucket, Aneroidbarometer
Regularsurfacemetparameters athourlyintervalSupplementarymeteorological data AutomaticWeatherSta- tionsAstraHourlyvaluesofapparentwind speed,direction,T,RH,rainfall andsunshine
Meteorologicaldata
Figure 3. (a)Details of the aircraft sorties during ICARB.
The lines from each base show the different sorties car- ried out from that base.(b)The instrumental setup in the aircraft.
The Micro Pulse Lidar (MPL) required no aspi- ration of ambient air and was to be operated from a high altitude of ∼8 km and as such, the aircraft was pressurized during this sortie. To avoid the large background and strong ground reflections, the MPL flight was restricted to early night-time.
The MPL (model MPL1000 of Science and Engi- neering Services Inc., USA) was fitted upside down in the aircraft. The MPL uses an AlGaAs diode pumped Nd-YLF laser, converting the primary radiation at 1047 nm to its second harmonic at 523.5 nm, at a pulse energy of 10μJ and a pulse repetition frequency (PRF) of 2.5 kHz. The col- lected data were analysed following the particulate- free zone approach and described in Satheesh et al (2006). The optical flat window of the aircraft meant for its survey operations provided the ideal window for lidar operations. The configuration of
the instruments within the aircraft is shown in figure 3(b).
4.4 Satellite component
For such a large field experiment covering vast areas of the land and ocean, it is imperative to have adequate complementary satellite data and derived products. For ICARB, the cloud pictures from METEOSAT and KALPANA satellites, and AODs derived from MODIS and NOAA-18 AVHRR were extensively used. Besides, extensive use of the NCEP/NCAR reanalysis data as well as the NOAA HYSPLIT air trajectory models were extensively used, both during the campaign for fine-tuning the cruise tracks and experimental schedules, as well as during the post-cruise data analysis.
5. Results
In the following, we present a few preliminary results of ICARB and evaluate them for the spatio- temporal heterogeneities and against the back- drop of earlier investigations in this region. More detailed results are provided in several companion papers in this issue, by the respective investigating teams.
5.1 Spatial distribution of aerosol optical depth over the Arabian Sea and BoB
The spectral AOD measurements, made onboard the ORV during its Arabian Sea leg (SK223B), are combined to examine the average spatial dis- tribution in figure 4. In doing so it is implicitly assumed that the spatial pattern remained statis- tically stationary during the rather short span of
∼24 days of the cruise. This is also justified by the earlier measurements over different parts of Arabian Sea during this season (e.g., Moorthy and Satheesh 2000; Moorthy et al 2005a). It is quite interesting to note from the figure that:
• Except at two pockets of moderate to high AODs, over most of the Arabian Sea the AODs are quite low, varying between 0.06 and 0.30 at 500 nm. In several parts of the central Arabian Sea, the values were between 0.06 and 0.12.
• There is one region, off the coast of Mangalore, where the AODs are very high, reaching as high as 0.85. This region of high AOD and high aerosol concentration appears to be a consistent feature, and was reported during several earlier cruises (Moorthy and Saha 2000; Parameswaran et al 1999; Ramachandran 2004). During a road campaign of 2004 winter, Moorthy et al (2005b) have reported high particulate mass
Table 4. Instruments operated onboard the aircraft.
Instruments Parameters measured
Aethalometer Black carbon mass concentration (MB)
Nephelometer Aerosol scattering coefficient at 450, 550 and 700 nm Scanning mobility particle sizer Aerosol number size distribution
Optical particle counter Aerosol mass size distribution Micro Pulse Lidar Aerosol backscatter and extinction
Ozone analyzer Ozone concentration
GPS receiver Instantaneous position of the aircraft
loading along the coast around Mangalore. Dur- ing the ICARB, this high is quite pronounced and strong.
• Another region, of smaller spatial extent, of moderately high AOD (0.3 to 0.4 at 500 nm) is seen centered at 18◦N; 68◦E. It might be recalled that based on the INDOEX data of March 1998 and 1999, Moorthy and Saha (2000) have reported the occurrence of a region of high AOD around the same location, which they called the West Asian High, mainly because of its proxim- ity to the west Asian region. In the same region, based on conductivity measurements, Kamra et al (2001) have reported a region of enhanced aerosol concentration. The high, observed in the present study, is in-line with the earlier findings, notwithstanding that it is much weaker than what was seen during the INDOEX. Besides this, there exists a very small region of moderate AOD off Mumbai.
• Over the rest of the Arabian Sea, the AOD is extremely low. It is also interesting to note that over the wide region bound between 56◦E and 68◦E; 9◦N and 15◦N, the AODs are low (∼0.15 at 500 nm). In this same region, indepen- dent measurements of surface mass concentra- tions (using a QCM impactor onboard the same cruise) also revealed extremely low mass concen- trations (<15μg m−3) as reported by Nairet al (this issue).
At this juncture it would be interesting to exam- ine the spatial distribution of AOD over the BoB.
The spatial composite of AOD at 500 nm wave- length, is shown in figure 5(a) using the measure- ments from the first leg of the ocean segment of the ICARB onboard cruise SK223A, even though these data were obtained during March–April 2006, about one month prior to the AS measurements. In general, higher AODs prevailed over the BoB, com- pared to AS. In contrast to the AS pattern, there is a sharp latitudinal gradient in the BoB, with the AODs increasing northward. Moderate-to-high AODs occur over the entire Head BoB, north of 15◦N and as far as 92◦E from the eastern coast of
Figure 4. Spatial distribution of AOD at 500 nm over the Arabian Sea during ICARB.
Indian mainland. In this region, AOD at 500 nm is in the range 0.5 to 1. Interestingly there occurs rather large region of approximately 3◦×3◦ in size and centered about 17.4◦N, 87.1◦E, where the AOD is extremely large, exceeding 0.8 and going as high as 1.1. We designate this as a ‘detached high’ due to its isolated nature, with no connection to the mainland. An examination of the prevailing winds (at 850 and 700 hPa levels from the NCEP reanalysis data) revealed that this region was under the influence of a low-level anticyclonic circula- tion, which spatially confines the aerosols. The back-trajectory analysis revealed strong advec- tion pathways from the northwest (west Asia and across the Indo Gangetic Plains). The spa- tial variation of AOD over the central Indian region during this period revealed strong impact of western advection (Beegum et al, this issue).
Aircraft profiles (analysis yet to be completed) off- Bhubaneswar also showed the presence of regions of high aerosol concentration above 1.2 km altitude (probably associated with advection). Using col- located measurements of AOD and altitude pro- files of aerosol extinction using a micro pulse lidar
Figure 5. (a)Spatial variation of AOD at 500 nm over the Bay of Bengal.(b)Latitudinal variation of AOD over BoB.
The error bars represent standard errors.
from Visakhapatnam, located on the east coast (figure 1), Niranjan et al (2007) have reported (episodic) occurrence of high-altitude aerosol lay- ers during the months of March/April 2005, 2006, in the lidar-derived profiles, which produced an increase in the AOD by 0.05 to 0.25. With the aid of HYSPLIT back-trajectory analysis they showed that the possible origin of these layers could be advection of mineral dust from Arabia in 60% of the cases, while it could be of the IGP origin dur- ing the rest of the events. Thus it appears that the combined effects of this advection and the low-level anticyclonic circulation over the BoB resulting in a spatial confinement of aerosols might have led to the formation of the detached high over the ocean.
In figure 5(b) we show the latitudinal variation of AOD at 500 nm, averaged over the longitudes 83◦E to 92◦E (so as to eliminate the values that are too close to the coast). It shows a gradual build- up from the lowest value (∼0.17) at 8◦N to reach the peak at 17◦N and then a weak decrease. Below 5◦N, a very weak increase is depicted, the reason for this is not clear except that there exists a narrow
Figure 6. Monthly mean regional distribution of AOD derived from NOAA-18-AVHRR data during February, March, and April 2006. The white shade shows regions where AOD could not be retrieved due to clouds (courtesy – K Rajeevet al, SPL).
shipping channel between 5 and 6◦N, which handles nearly 25% of the international ship traffic.
5.2 Regional distribution of AOD during February–April 2006 from satellite data In view of the above observations, which revealed significant differences from the earlier reports, the monthly mean spatial distribution of AODs (at the wavelength 630±50 nm) over the AS, BoB, and equatorial Indian Ocean (EIO) were examined. These AODs were retrieved from the Advanced Very High Resolution Radiometer (AVHRR) onboard NOAA-18 satellite (Level-1b Global Area Coverage data of the afternoon passes) using the operational satellite sensor cali- bration coefficients following the details outlined in Rajeev et al (2000). The results for three months, February, March, and April 2006, are shown in fig- ure 6. The significant observation from figure 5 is the general conformity to the cruise observa- tions during the month of April, where the high AODs (>0.5 at 630 nm) off Mangalore coast and the low values (0.1 to 0.2) in the central Arabian
Sea (bound between 55◦E to 65◦E; and equator to 15◦N) are clearly seen in the bottom panel. Look- ing back to the temporal evolution of AOD, the upper panels of the same figure show that:
• In February, two distinct regions of high AOD values are observed over BoB, one in the north and the other in the east off the Myanmar coast.
Over most parts of the Arabian Sea, AOD is between 0.2 and 0.3 with relatively smaller spa- tial variation.
• The AODs significantly decrease from February to March, over the entire Arabian Sea and BoB.
Over the Arabian Sea, except very near the con- tinent, AOD is generally less than 0.2 in March and very low values are encountered particularly over the central and eastern parts.
Comparing these observations with the mean regional distribution of AOD obtained from 7 years’ data over this region (Nair et al 2005), the major difference is that while the AOD distri- bution shows an increase over the eastern Arabian Sea from February to March in the mean pattern, it shows a decrease in the year 2006.
5.3 Spatial variation of ˚Angstr¨om parameters over the Arabian Sea during ICARB By performing a regression analysis of the spec- tral AODs with the ˚Angstr¨om relation (˚Angstr¨om 1964) τpλ =βλ−α (whereτpλ is the AOD at wave- length λexpressed in micrometer) the wavelength exponent α and turbidity coefficient β are eval- uated. While the wavelength exponent is a mea- sure of the ratio of accumulation mode to coarse mode concentrations of the columnar aerosols (and hence an indicator of the aerosol columnar size dis- tribution), the coefficient β is a measure of the column abundance of particles. In figure 7(a) we show the spatial variation of α over the Arabian Sea while that ofβ is examined in figure 7(b). The most striking features are:
• The southern Arabian Sea (south of 15◦N) is characterized by high values of α (and hence steep AOD spectra) with values in the range 0.85 to 1.1, indicating high relative abundance of accumulation mode aerosols. Compared to this, the northern Arabian Sea is dominated by coarser mode aerosols and α goes well below 0.6. These coarse mode aerosols could be asso- ciated with either sea spray aerosols or trans- ported mineral dust or both.
• Over most of the Arabian Sea β remains very low (below 0.15), showing low columnar aerosol loading except for the two pockets, where high AODs were seen in figure 4. However, off Mangalore coast, we note a pocket of enhanced
Figure 7. (a) Spatial distribution of the ˚Angstr¨om expo- nent over the Arabian Sea during ICARB.(b)Spatial dis- tribution of the turbidity coefficient over Arabian Sea.
abundance of accumulation mode particles (high AOD, highα, and highβ). On the other hand at the second pocket in the central AS where AOD was moderate, even though β also was moder- ate, α was low implying that the concentration of accumulation mode aerosols was low. It is also important to note that these findings are consis- tent with the mass concentration measurements of aerosols near the surface with the QCM (Nair et al, this issue), which showed higher accumula- tion fraction near the coast and in the southern Arabian Sea, while in the northern AS the coarse mode fraction was higher.
5.4 AOD distribution over mainland and islands
Spectral AODs, obtained regularly during the ICARB period from the network observatories spread over the Indian mainland and adjoin- ing islands (figure 1), are averaged over the months and the spatial composites are shown in
